Information for corpus elves

• Basic information
• Transcription
• Part-of-speech tagging
• Syntactic annotation
• Further resources
• Who's working on what

• Password security
• First steps in using your babel account
• Basics of Linux/Unix
• Basics of Emacs (for lists of commands, google "emacs cheat sheet")

• Entering accented characters
• French transcription conventions

• Lexicon mode reference
• Ottawa POS annotation manual (see also Penn syntactic annotation manual)
  • In contrast to the Ottawa tagging conventions, the Penn texts distinguish clitic pronouns (CL) from ordinary pronouns (PRO). For more information, see CL versus PRO.
• Mots difficiles (Penn version)
• Division into sentence tokens
  • The English annotation system uses period and comma as POS tags, where the French annotation system uses PONFP and PON, respectively. But this difference doesn't affect the criteria for dividing tokens, which are the same for both languages.
  • If a sentence token ends without punctuation, add an <+ EOS +> comment as you're correcting the POS tags. The comment will be converted into a fake punctuation mark that helps the parser, after which it is deleted.

• Penn syntactic annotation manual
• When using CorpusDraw, in order for things to work out well, you need to invoke it from a different directory than the one containing the parsed file that you are correcting. Usually, what people do is keep the files they're correcting in a parsed directory and invoke CorpusDraw from that directory's mother directory. So, to invoke CorpusDraw on .../working-dir/parsed/my-file.psd, you'd cd to working-dir and say

      CD parsed/my-file.psd
      

• When the CorpusDraw window opens up, it's a good idea to use the "Zoom Window" option under the "Window" tab in the top bar (the one associated with X11). If you don't zoom the window, the scroll bars tend not to work correctly.
• CorpusDraw keyboard shortcuts
• When CorpusDraw saves a file that you're working on, it saves it with an additional .new extension. So my-file.psd after saving becomes my-file.psd.new. This means that any time you quit CorpusDraw, you need to update your file names. The original file with the .psd extension should become your backup file with an additional .old extension, and the .psd.new file becomes the current .psd file. You can do this automatically using the update command from the directory just above your parsed directory (that is, the directory from which you invoke CorpusDraw). The argument of the update command should be just the basename of the file that you're working on (no need to include the parsed subdirectory or any extensions to the filename). The command automatically Does The Right Thing. For example:
  

      update my-file
      

  turns .../parsed/my-file.psd into .../parsed/my-file.psd.old, and .../parsed/my-file.psd.new into .../parsed/my-file.psd.

• Help with grep and regular expressions
• Anglo-Norman dictionary
  select "Browse or search the Anglo-Norman dictionary"
• Middle French dictionary (Godefroy)
• Philological information concerning the texts in the French corpus (under construction)
• Vulgate (with English translation)

/home/migration/other/MIDENG/MIDDLE-FRENCH/tagged-uncorrected/

cd tagging
cp /home/migration/other/MIDENG/MIDDLE-FRENCH/tagged-uncorrected/ot-genesis-08-to-14.lex .